Current network is overburdened, and requires optimization, due to continuous enlargement of network, continuous growth of data traffic and complexity of service.
Traffic Engineering (TE) is developed as required, which focuses on optimization of general performance of networks, so as to provide efficient and stable network services, optimize usage of network resources, and optimize network traffic. In terms of difference of functionality ranges, the traffic engineering can be categorized into intra-domain traffic engineering and inter-domain traffic engineering. At present, the intra-domain traffic engineering is relatively mature, the major method thereof is to flood TE information of links in the intra-domain via extension of conventional interior gateway protocol (IGP), such as extension of Open Shortest Path First Traffic Engineering (OSPF-TE), and extension of Intermediate System to Intermediate System Traffic Engineering (ISIS-TE), so as to implement synchronization of intra-domain Traffic Engine Database (TED).
In a typically large network including a plurality of Autonomous Systems, in consideration of extensibility and security, the TE information within each AS cannot be flooded to the outside of the autonomous system. In order to obtain TE path between the autonomous systems, there are mainly two path calculation methods: one method is to utilize a Path Computation Element (PCE) to cooperatively calculate, i.e. within the autonomous systems there is one or more PCEs responsible for the path calculation of the present autonomous system, and the path calculation between the autonomous systems is completed by the PCE cooperation with one or more PCEs of other autonomous systems; another method is to utilize the manner of explicit route to display AS sequence of a specified path or IP address sequence of a specified Autonomous System Border Router (ASBR), and then complete the path calculation among the autonomous systems according to the specified information by a Label Switched Router (LSR). No matter which calculation method is utilized, the Path Computation Element (PCE) or the Label Switched Router (LSR) responsible for TE path calculation needs to know border connection information between the autonomous system located therein and external neighbor autonomous systems. The border connection information refers to information of establishing or deleting connection between Autonomous System Border Router (ASBR) where the path calculating node is located and the external neighbor Autonomous System Border Router (ASBR).
Currently, no solution raises how to advertise border connection information of autonomous systems. The border connection information is statically and manually configured as required by the nodes within the systems. However, the disadvantage of the static and manual configuration lies in that it fails to adapt to dynamic change of network topology, so that it cannot support the TE path calculation between autonomous systems efficiently, and then it is difficult to implement optimization of general performance of networks. When the border connection information of the autonomous systems changes, such as a disconnection between a certain border node and a corresponding neighbor border node in a neighbor autonomous system, if the node responsible for path calculation within the autonomous system obtains the border connection information of the present system via a static and manual configuration method, it always fails to learn above-mentioned change of border connection information in time, and executes the path calculation based on the previous information statically configured. The path may be noticed unavailable during the calculating process and need to be re-configured via a manual modification. Additionally, when the border connection relationship of autonomous systems is relatively complex, the static and manual configuration method has problems, such as high workload, time consuming, low processing efficiency and high cost, and wrong-match or miss-match may easily occur because of human cause.